Skip to main content

Intercom

The Intercom agent connector is a Python package that equips AI agents to interact with Intercom through strongly typed, well-documented tools. It's ready to use directly in your Python app, in an agent framework, or exposed through an MCP.

Intercom is a customer messaging platform that enables businesses to communicate with customers through chat, email, and in-app messaging. This connector provides access to core Intercom entities including contacts, conversations, companies, teams, admins, tags, and segments for customer support analytics and insights. It supports creating, updating, and deleting contacts and companies; creating and deleting tags; creating, updating, and deleting conversations; creating notes; and creating, updating, and deleting internal articles.

Example prompts​

The Intercom connector is optimized to handle prompts like these.

  • List all contacts in my Intercom workspace
  • List all companies in Intercom
  • What teams are configured in my workspace?
  • Show me all admins in my Intercom account
  • List all tags used in Intercom
  • Show me all customer segments
  • Show me details for a recent contact
  • Show me details for a recent company
  • Show me details for a recent conversation
  • Create a new lead contact named 'Jane Smith' with email jane@example.com
  • Create an internal article titled 'Onboarding Guide' with instructions for new team members
  • Create a company named 'Acme Corp' with company_id 'acme-001'
  • Create a tag named 'VIP Customer'
  • Create a conversation from contact {id} saying 'I need help with my account'
  • Update the name of contact {id} to 'John Updated'
  • Add a note to contact {id} saying 'Followed up on support request'
  • Show me conversations from the last week
  • List conversations assigned to team {team_id}
  • Show me open conversations
  • Delete contact {id}
  • Delete company {id}
  • Delete tag {id}
  • Delete conversation {id}
  • Delete internal article {id}
  • Update conversation {id} to mark it as read
  • Update internal article {id} with a new title

Unsupported prompts​

The Intercom connector isn't currently able to handle prompts like these.

  • Send a message to a customer
  • Assign a conversation to an admin

Entities and actions​

This connector supports the following entities and actions. For more details, see this connector's full reference documentation.

EntityActions
ContactsList, Create, Get, Update, Delete, Context Store Search, Context Store SQL Query
ConversationsList, Create, Get, Update, Delete, Context Store Search, Context Store SQL Query, Semantic Search
CompaniesList, Create, Get, Update, Delete, Context Store Search, Context Store SQL Query
TeamsList, Get, Context Store Search, Context Store SQL Query
AdminsList, Get
TagsList, Create, Get, Delete
NotesCreate
SegmentsList, Get
Internal ArticlesCreate, Update, Delete

Intercom API docs​

See the official Intercom API reference.

Interfaces​

Use the Intercom connector through the Airbyte Agent CLI, the Python SDK, or the API.

CLI​

Install the CLI:

curl -fsSL https://airbyte.ai/install.sh | bash

Authenticate with Airbyte:

airbyte-agent login

Create the connector. The CLI opens the hosted setup flow:

airbyte-agent connectors create --json '{
  "workspace": "<your_workspace_name>",
  "name": "intercom"
}'

Describe the connector to see its supported entities and actions:

airbyte-agent connectors describe --json '{
  "workspace": "<your_workspace_name>",
  "name": "intercom"
}'

Execute an action:

airbyte-agent connectors execute --json '{
  "workspace": "<your_workspace_name>",
  "name": "intercom",
  "entity": "contacts",
  "action": "list"
}'

Python SDK​

Installation​

uv pip install airbyte-agent-sdk

Usage​

Connectors can run in hosted or open source mode.

Hosted​

In hosted mode, API credentials are stored securely in Airbyte Agents. You provide your Airbyte credentials instead. If your Airbyte client can access multiple organizations, also set organization_id.

This example assumes you've already authenticated your connector with Airbyte. See Authentication to learn more about authenticating. If you need a step-by-step guide, see the hosted execution tutorial.

The connect() factory returns a fully typed IntercomConnector and reads AIRBYTE_CLIENT_ID / AIRBYTE_CLIENT_SECRET from the environment:

The recommended pattern is build_connector_tools, which gives the agent three tools bound to this connector: inspect_connector, read_skill_docs, and execute. The agent can inspect the connector, read only the skill-doc section it needs, and then execute:

inspect_connector() -> read_skill_docs() -> read_skill_docs(section="...") -> execute(entity, action, params)

Pass section IDs verbatim as the outline lists them, prefix included (actions.<entity>.<action>, not <entity>.<action>); anything else returns an error the agent has to recover from.

The builder names its tools inspect_connector, read_skill_docs, and execute, so the tool sets for more than one connector collide when registered on the same agent. Renaming the callables at registration avoids the collision, but the generated execute guidance still names inspect_connector and read_skill_docs, pointing the model at the wrong tools. Use the agent_tool pattern below instead: it weaves your own names into that guidance.

Pydantic AI
from airbyte_agent_sdk import build_connector_tools
from pydantic_ai import Agent
from airbyte_agent_sdk import connect
from airbyte_agent_sdk.connectors.intercom import IntercomConnector

connector = connect("intercom", workspace_name="<your_workspace_name>")

tools = build_connector_tools(connector, framework="pydantic_ai")
agent = Agent("openai:gpt-4o", tools=tools.as_list())
Custom tool bodies​

When you need custom tool bodies — or a framework without native support — use IntercomConnector.agent_tool. Register execute, inspect, and docs together so the agent can fetch connector guidance progressively. Pass the framework explicitly when it has a supported failure strategy:

Pydantic AI
from pydantic_ai import Agent
from airbyte_agent_sdk import connect
from airbyte_agent_sdk.connectors.intercom import IntercomConnector

connector = connect("intercom", workspace_name="<your_workspace_name>")

agent = Agent("openai:gpt-4o")

@agent.tool_plain
@IntercomConnector.agent_tool(
    framework="pydantic_ai",
    inspect_tool="intercom_inspect",
    docs_tool="intercom_read_docs",
)
async def intercom_execute(entity: str, action: str, params: dict | None = None):
    return await connector.execute(entity, action, params or {})

@agent.tool_plain
@IntercomConnector.agent_tool(framework="pydantic_ai")
async def intercom_inspect():
    return await connector.inspect_connector()

@agent.tool_plain
@IntercomConnector.agent_tool(framework="pydantic_ai")
async def intercom_read_docs(section: str | None = None):
    return await connector.read_skill_docs(section)

Use the same three-function pattern with framework="langchain", "openai_agents", or "mcp" and that framework's registration decorator. Each value translates connector failures into the framework's own signal:

framework=Tool failures surface as
"pydantic_ai"pydantic_ai.ModelRetry
"langchain"langchain_core.tools.ToolException (set handle_tool_error=True to feed it back to the model)
"openai_agents"the failure message returned to the model as the tool result
"mcp"fastmcp.exceptions.ToolError
"none" (default)airbyte_agent_sdk.AirbyteToolError

On a framework the SDK does not support natively — or in a raw LLM dispatch loop — omit framework= and handle AirbyteToolError yourself:

No framework
from airbyte_agent_sdk import AirbyteToolError
from airbyte_agent_sdk import connect
from airbyte_agent_sdk.connectors.intercom import IntercomConnector

connector = connect("intercom", workspace_name="<your_workspace_name>")

@IntercomConnector.agent_tool(
    inspect_tool="intercom_inspect",
    docs_tool="intercom_read_docs",
)
async def intercom_execute(entity: str, action: str, params: dict | None = None):
    return await connector.execute(entity, action, params or {})

@IntercomConnector.agent_tool()
async def intercom_inspect():
    return await connector.inspect_connector()

@IntercomConnector.agent_tool()
async def intercom_read_docs(section: str | None = None):
    return await connector.read_skill_docs(section)

# Advertise all three to the model, using each function's docstring as its description.
handlers = {
    fn.__name__: fn
    for fn in (intercom_inspect, intercom_read_docs, intercom_execute)
}

# `tool_name` and `tool_args` come from the model's tool call in your dispatch loop.
try:
    tool_result = await handlers[tool_name](**tool_args)
except AirbyteToolError as err:
    tool_result = str(err)  # hand the message back to the model as an errored tool result

Each function's docstring carries the guidance the model needs, so pass it through as the tool description wherever you register it.

Legacy alternatives​

These examples are kept for existing integrations. The deprecated IntercomConnector.tool_utils pattern loads the connector's full generated catalog into one broad execute tool description instead of letting the agent read skill docs on demand. For new code, use build_connector_tools or IntercomConnector.agent_tool above.

Pydantic AI
from pydantic_ai import Agent
from airbyte_agent_sdk import connect
from airbyte_agent_sdk.connectors.intercom import IntercomConnector

connector = connect("intercom", workspace_name="<your_workspace_name>")

agent = Agent("openai:gpt-4o")

@agent.tool_plain
@IntercomConnector.tool_utils
async def intercom_execute(entity: str, action: str, params: dict | None = None):
    return await connector.execute(entity, action, params or {})

Or pass credentials explicitly (equivalent, useful when you're not loading them from the environment):

Pydantic AI
from airbyte_agent_sdk import build_connector_tools
from pydantic_ai import Agent
from airbyte_agent_sdk.connectors.intercom import IntercomConnector
from airbyte_agent_sdk.types import AirbyteAuthConfig

connector = IntercomConnector(
    auth_config=AirbyteAuthConfig(
        workspace_name="<your_workspace_name>",
        organization_id="<your_organization_id>",  # Optional for multi-org clients
        airbyte_client_id="<your-client-id>",
        airbyte_client_secret="<your-client-secret>"
    )
)

tools = build_connector_tools(connector, framework="pydantic_ai")
agent = Agent("openai:gpt-4o", tools=tools.as_list())
Open source​

In open source mode, you provide API credentials directly to the connector.

The recommended pattern is build_connector_tools, which gives the agent three tools bound to this connector: inspect_connector, read_skill_docs, and execute. The agent can inspect the connector, read only the skill-doc section it needs, and then execute:

inspect_connector() -> read_skill_docs() -> read_skill_docs(section="...") -> execute(entity, action, params)

Pass section IDs verbatim as the outline lists them, prefix included (actions.<entity>.<action>, not <entity>.<action>); anything else returns an error the agent has to recover from.

The builder names its tools inspect_connector, read_skill_docs, and execute, so the tool sets for more than one connector collide when registered on the same agent. Renaming the callables at registration avoids the collision, but the generated execute guidance still names inspect_connector and read_skill_docs, pointing the model at the wrong tools. Use the agent_tool pattern below instead: it weaves your own names into that guidance.

Pydantic AI
from airbyte_agent_sdk import build_connector_tools
from pydantic_ai import Agent
from airbyte_agent_sdk.connectors.intercom import IntercomConnector
from airbyte_agent_sdk.connectors.intercom.models import IntercomAuthConfig

connector = IntercomConnector(
    auth_config=IntercomAuthConfig(
        access_token="<Your Intercom API Access Token>"
    )
)

tools = build_connector_tools(connector, framework="pydantic_ai")
agent = Agent("openai:gpt-4o", tools=tools.as_list())
Custom tool bodies​

When you need custom tool bodies — or a framework without native support — use IntercomConnector.agent_tool. Register execute, inspect, and docs together so the agent can fetch connector guidance progressively. Pass the framework explicitly when it has a supported failure strategy:

Pydantic AI
from pydantic_ai import Agent
from airbyte_agent_sdk.connectors.intercom import IntercomConnector
from airbyte_agent_sdk.connectors.intercom.models import IntercomAuthConfig

connector = IntercomConnector(
    auth_config=IntercomAuthConfig(
        access_token="<Your Intercom API Access Token>"
    )
)

agent = Agent("openai:gpt-4o")

@agent.tool_plain
@IntercomConnector.agent_tool(
    framework="pydantic_ai",
    inspect_tool="intercom_inspect",
    docs_tool="intercom_read_docs",
)
async def intercom_execute(entity: str, action: str, params: dict | None = None):
    return await connector.execute(entity, action, params or {})

@agent.tool_plain
@IntercomConnector.agent_tool(framework="pydantic_ai")
async def intercom_inspect():
    return await connector.inspect_connector()

@agent.tool_plain
@IntercomConnector.agent_tool(framework="pydantic_ai")
async def intercom_read_docs(section: str | None = None):
    return await connector.read_skill_docs(section)

Use the same three-function pattern with framework="langchain", "openai_agents", or "mcp" and that framework's registration decorator. Each value translates connector failures into the framework's own signal:

framework=Tool failures surface as
"pydantic_ai"pydantic_ai.ModelRetry
"langchain"langchain_core.tools.ToolException (set handle_tool_error=True to feed it back to the model)
"openai_agents"the failure message returned to the model as the tool result
"mcp"fastmcp.exceptions.ToolError
"none" (default)airbyte_agent_sdk.AirbyteToolError

On a framework the SDK does not support natively — or in a raw LLM dispatch loop — omit framework= and handle AirbyteToolError yourself:

No framework
from airbyte_agent_sdk import AirbyteToolError
from airbyte_agent_sdk.connectors.intercom import IntercomConnector
from airbyte_agent_sdk.connectors.intercom.models import IntercomAuthConfig

connector = IntercomConnector(
    auth_config=IntercomAuthConfig(
        access_token="<Your Intercom API Access Token>"
    )
)

@IntercomConnector.agent_tool(
    inspect_tool="intercom_inspect",
    docs_tool="intercom_read_docs",
)
async def intercom_execute(entity: str, action: str, params: dict | None = None):
    return await connector.execute(entity, action, params or {})

@IntercomConnector.agent_tool()
async def intercom_inspect():
    return await connector.inspect_connector()

@IntercomConnector.agent_tool()
async def intercom_read_docs(section: str | None = None):
    return await connector.read_skill_docs(section)

# Advertise all three to the model, using each function's docstring as its description.
handlers = {
    fn.__name__: fn
    for fn in (intercom_inspect, intercom_read_docs, intercom_execute)
}

# `tool_name` and `tool_args` come from the model's tool call in your dispatch loop.
try:
    tool_result = await handlers[tool_name](**tool_args)
except AirbyteToolError as err:
    tool_result = str(err)  # hand the message back to the model as an errored tool result

Each function's docstring carries the guidance the model needs, so pass it through as the tool description wherever you register it.

Legacy alternatives​

These examples are kept for existing integrations. The deprecated IntercomConnector.tool_utils pattern loads the connector's full generated catalog into one broad execute tool description instead of letting the agent read skill docs on demand. For new code, use build_connector_tools or IntercomConnector.agent_tool above.

Pydantic AI
from pydantic_ai import Agent
from airbyte_agent_sdk.connectors.intercom import IntercomConnector
from airbyte_agent_sdk.connectors.intercom.models import IntercomAuthConfig

connector = IntercomConnector(
    auth_config=IntercomAuthConfig(
        access_token="<Your Intercom API Access Token>"
    )
)

agent = Agent("openai:gpt-4o")

@agent.tool_plain
@IntercomConnector.tool_utils
async def intercom_execute(entity: str, action: str, params: dict | None = None):
    return await connector.execute(entity, action, params or {})

Authentication​

For all authentication options, see the connector's authentication documentation.

IP allow list​

If your organization restricts access to specific IPs, add the Airbyte Agents IP addresses to your allow list.

Version information​

Connector version: 0.1.10

Was this page helpful?